I have a pair of Electra-Print SE outputs that I had to re-stack after some rough shipping & "handling"....
They were not "dipped" - impregnated. Therefore the lams shifted when they were dropped. The paper gapping material was ripped and not flat anymore.
Is there a formula for determining the thickness? I have used some similar paper to re-gap them but I am wondering how critical it is?
They were supposedly for a GM-70, gapped for 125ma with priority given to HF response. They have perhaps only a quarter of the winding window on the bobbin filled. The lams are about the same size as Hammond 1628SEA. And it is only P-S-P layered. Is this normal? I though Electra-print had a good reputation for quality, but scramble wound with no interleaving really is something I could do myself... What am I missing?
Thanks for the help.
They were not "dipped" - impregnated. Therefore the lams shifted when they were dropped. The paper gapping material was ripped and not flat anymore.
Is there a formula for determining the thickness? I have used some similar paper to re-gap them but I am wondering how critical it is?
They were supposedly for a GM-70, gapped for 125ma with priority given to HF response. They have perhaps only a quarter of the winding window on the bobbin filled. The lams are about the same size as Hammond 1628SEA. And it is only P-S-P layered. Is this normal? I though Electra-print had a good reputation for quality, but scramble wound with no interleaving really is something I could do myself... What am I missing?
Thanks for the help.
Nobody else - only the original manufacturer - can tell you the exact air gap requirement for a particular installation. The simplest would be to ask them for the info. You can determine the airgap if you have measurement possibilities. You need to measure the inductance with DC current excitation. With the right airgap the inductance curve is flat up to the dc current of the tube. This type of instrument is very scarce. Another method is described in Sound Practices, unfortunately i don't find it right now, which needs more multimeters and a power rheostat. The title of the article is something as "Use your old junkbox chokes" or similar. Do not put a normal piece of paper in the transformer as air gap!!! Most of the papers contain very corrosive substances, make the iron rusty. Use such materials which are used in transformer windings as electrical row or layer insulation.
Thnaks for the reply. I had planned on using some "Back-Papier" its a type of food baking paper and is very sturdy and non toxic. I measured that 2 pieces stacked is almost exactly the same width as the original.
If I reduced the gap slightly would I gain bass response with a lower MADC flowing through the trans. due to the increased inductance?
If I reduced the gap slightly would I gain bass response with a lower MADC flowing through the trans. due to the increased inductance?
I thought the air gap only affected the low frequency response (inductance), DC saturation, and Low Frequency Power saturation.
I think high frequency optimization is more dependent on the windings, layers, wire size, etc.
If I remember correctly, Electra Print Audio checks the air gap setting by running the transformer in an actual amplifier, under DC current, and Low frequency current (signal).
At one time, I used Kapton Tape to fill the Air Gap, one or two layers. I can not speak as to whether the tape and/or adhesive may or may not corrode the laminations. When I mentioned using the Kapton tape to others, nobody objected.
I think high frequency optimization is more dependent on the windings, layers, wire size, etc.
If I remember correctly, Electra Print Audio checks the air gap setting by running the transformer in an actual amplifier, under DC current, and Low frequency current (signal).
At one time, I used Kapton Tape to fill the Air Gap, one or two layers. I can not speak as to whether the tape and/or adhesive may or may not corrode the laminations. When I mentioned using the Kapton tape to others, nobody objected.
Last edited:
Theoretically you can have more bass by reducing the airgap if the original design had had some margin. Reducing the idle current is a bit more complicated issue. On one hand it makes possible to decrease the air gap, but on the other hand increase the tube internal resistance and decrease the generator capability to drive the transformer which decreases the bass response. There are two ways to proceed, first: measurements and measurements. Second: listening tests - trial and error. The latter way can be very misleading.
BarG01,
Good points.
Generally: If you parallel two tubes, even though the current in each one is reduced to accommodate the transformer DC current, You may either get more total quiescent current to get a lower parallel resistance, or you may not get lower parallel plate resistance than the original single tube at its original 'full' quiescent current. Generally: For more DC capability, and for more Low Frequency Power, we need more core.
I believe if the Electra Print transformer has a re-set air gap that is reasonably close to the original spacing, it will perform pretty close to what it did when it left the manufacturer.
I have used SE and push pull Electra Print transformers, as well as other manufacturers.
Good points.
Generally: If you parallel two tubes, even though the current in each one is reduced to accommodate the transformer DC current, You may either get more total quiescent current to get a lower parallel resistance, or you may not get lower parallel plate resistance than the original single tube at its original 'full' quiescent current. Generally: For more DC capability, and for more Low Frequency Power, we need more core.
I believe if the Electra Print transformer has a re-set air gap that is reasonably close to the original spacing, it will perform pretty close to what it did when it left the manufacturer.
I have used SE and push pull Electra Print transformers, as well as other manufacturers.
Last edited:
You are right. If we restore the original airgap, that should work.
But, if we want to get to the edge, we can, if the original design comprised some margin. Designers always build in some margin in their designs to avoid later problems. The question is that how much margin is built in this particular product? To discover this, two ways mentioned above may be used.
But, if we want to get to the edge, we can, if the original design comprised some margin. Designers always build in some margin in their designs to avoid later problems. The question is that how much margin is built in this particular product? To discover this, two ways mentioned above may be used.
There is an easy way to adjust the air gap. You need an oscilloscope, and audiofrequency oscillator and a resistive load. Inject a signal of about 40Hz (or the lowest expected signal thought it) into the amp loaded with nominal resistive load. Increase the signal until the maximum expected power ouput is taken from it. If the amp has NFB from the secondary of the traffo, then remove it temporarily. Carefully, increase or decrease the gap until get the minimum distortion of the signal at the load resistance.
This is because the trafo will saturate if too low airgap is set into the core, and if the gap is too large, inductance decreases quickly. Then, there is a point at the DC normal current that distortion is at a minimum. I used this method not only to adjust air gaps in transformers whose primary includes DC current, also with chokes for power supplies.
This is because the trafo will saturate if too low airgap is set into the core, and if the gap is too large, inductance decreases quickly. Then, there is a point at the DC normal current that distortion is at a minimum. I used this method not only to adjust air gaps in transformers whose primary includes DC current, also with chokes for power supplies.
Attachments
Member
Joined 2009
Paid Member
For output transformers, I believe the primary inductance is usually rated in Henrys. I believe the leakage inductance is usually rated in milli Henrys. They need to be orders of magnitude apart, or you do not get good bandwidth.
I do not think the air gap influences the leakage reactance.
But it may very slightly affect the resonance, due to a small change in capacitance. No gap at all will cause the most capacitance from the windings to the I laminations.
Primary capacitance to I
Secondary capacitance to I
Therefore, it is also a small part of the capacitance from Primary to Secondary (via the I laminations).
A very large air gap would reduce the capacitance from the windings to the I laminations.
I think we are starting to get way into the inside layers of the onion with this effect that so much less than the other effects of transformer geometry. Lets peel the more significant outside layers off first.
I do not think the air gap influences the leakage reactance.
But it may very slightly affect the resonance, due to a small change in capacitance. No gap at all will cause the most capacitance from the windings to the I laminations.
Primary capacitance to I
Secondary capacitance to I
Therefore, it is also a small part of the capacitance from Primary to Secondary (via the I laminations).
A very large air gap would reduce the capacitance from the windings to the I laminations.
I think we are starting to get way into the inside layers of the onion with this effect that so much less than the other effects of transformer geometry. Lets peel the more significant outside layers off first.
I read this dependence in a professional book by Gerhard H. Domsch and my experiencies backed it.
How we measure the leakage inductance? By shorting the secondary we measure the inductance of the primary winding. If the leakage inductance were dependent only on the geometrical properties of the windings, sizes and distances, varying with the iron core would be completely ineffective to the leakage inductance. However the leakage inductance will change significantly with these modifications.
How we measure the leakage inductance? By shorting the secondary we measure the inductance of the primary winding. If the leakage inductance were dependent only on the geometrical properties of the windings, sizes and distances, varying with the iron core would be completely ineffective to the leakage inductance. However the leakage inductance will change significantly with these modifications.
Uh, . . . I believe the following is true about leakage inductance:
When you short the secondary, you are effectively shorting all of the main (very large) inductance of the primary. The only inductance that is left that you are measuring IS the leakage inductance in series with the primary DCR.
And, leakage inductance can be measured both ways.
You can also short the primary, and measure the secondary. You are effectively shorting all of the main (medium) inductance of the secondary. All that is left IS the leakage inductance in series with the secondary DCR.
In both cases you get the leakage inductance. But one is referred to the primary, and the other is referred to the secondary.
You will find some manufacturers list the leakage inductance referred to the primary. Other manufacturers list the leakage inductance referred to the secondary. And some manufacturers are either afraid, or unable to measure, or reducing cost . . . so they do not list the leakage incuctance at all.
Practical Measurements: I used a Rohde & Schwarz $50,000 Vector Network Analyzer (VNA). It had very specialized inputs and outputs, and $5,000 cal kits (open, short, load) that had cal factors for inductance, capacitance, and resistance of each piece in the kit. I also had a specialized bridge that allowed me to measure those audio output and interstage transformers from 10 Hz to many MHz.
The Rohde & Schwarz VNA could also measure bandwidth, phase, distributed capacitance, resonance, Q, insertion loss, etc.
I even measured the frequency response and phase versus frequency of non-feedback amplifiers into an 8 Ohm load, and then measured the frequency response and phase versus frequency of non-feedback amplifiers into various loudspeakers. Then I compared the two sets of frequency response and phase versus frequency measurements (resistive and loudspeaker loads).
Those were fun days when I had access to the $$$ Rohde & Schwarz VNA to use for all those measurements.
When you short the secondary, you are effectively shorting all of the main (very large) inductance of the primary. The only inductance that is left that you are measuring IS the leakage inductance in series with the primary DCR.
And, leakage inductance can be measured both ways.
You can also short the primary, and measure the secondary. You are effectively shorting all of the main (medium) inductance of the secondary. All that is left IS the leakage inductance in series with the secondary DCR.
In both cases you get the leakage inductance. But one is referred to the primary, and the other is referred to the secondary.
You will find some manufacturers list the leakage inductance referred to the primary. Other manufacturers list the leakage inductance referred to the secondary. And some manufacturers are either afraid, or unable to measure, or reducing cost . . . so they do not list the leakage incuctance at all.
Practical Measurements: I used a Rohde & Schwarz $50,000 Vector Network Analyzer (VNA). It had very specialized inputs and outputs, and $5,000 cal kits (open, short, load) that had cal factors for inductance, capacitance, and resistance of each piece in the kit. I also had a specialized bridge that allowed me to measure those audio output and interstage transformers from 10 Hz to many MHz.
The Rohde & Schwarz VNA could also measure bandwidth, phase, distributed capacitance, resonance, Q, insertion loss, etc.
I even measured the frequency response and phase versus frequency of non-feedback amplifiers into an 8 Ohm load, and then measured the frequency response and phase versus frequency of non-feedback amplifiers into various loudspeakers. Then I compared the two sets of frequency response and phase versus frequency measurements (resistive and loudspeaker loads).
Those were fun days when I had access to the $$$ Rohde & Schwarz VNA to use for all those measurements.
Last edited:
